Information Technology Reference
In-Depth Information
Building Castles in the Air
In describing his experiences as the project manager of the single-largest
and most expensive software development effort ever undertaken in the
history of the IBM Corporation, the noted computer scientist Frederick
Brooks provided a curiously literary portrayal of the computer program-
mer: “The programmer, like the poet, works only slightly removed from
pure-thought stuff. He builds his castles in the air, from air, creating by
exertion of the imagination.”
38
That a technical manager in a conservative corporation should use
such lofty language in reference to such a seemingly prosaic occupation
like programming is striking yet not unusual. But Brooks meant his liter-
ary metaphors to be taken seriously. Even more so than the poet, he
argued, the programmer worked in the medium of the imagination, using
words to bring to life grand conceptual structures. In fact, in the case of
the programmers the relationship between words and reality was almost
magical: “One types the correct incantation on a keyboard, and a display
screen comes to life, showing things that never were nor could be.” And
like the magical incantation, the computer program demanded perfec-
tion: “If one character, one pause, of the incantation is not strictly in
proper form, the magic doesn't work.” This is what made programming
so diffi cult, he suggested: “Human beings are not accustomed to being
perfect, and few areas of human activity demand it. Adjusting to the
requirement for perfection is, I think, the most diffi cult part of learning
to program.”
39
Like many of his contemporaries, Brooks was struggling to under-
stand why software development projects seemed almost impossible
to manage using conventional management techniques. In the late
1960s, Brooks had been the manager of the IBM OS/360 development
project. The OS/360 operating system was the cornerstone of IBM's
larger System/360 strategy, which consolidated IBM's computer product
lines into a single range of hardware- and software-compatible machines.
Although the System/360 turned out to be a tremendous success for
IBM, it had almost been derailed by problems with the development
of OS/360. In the years between 1963 and 1966, over fi ve thousand
staff years of effort went into the design, construction, and documenta-
tion of OS/360. When it was fi nally delivered in 1967, nine months
late and riddled with errors, it had cost the IBM Corporation half a
billion dollars—four times the original budget, and the single-largest
expenditure in company history. And according to Brooks, the personal
